Sampling Sufficiency for Fish Assemblage Surveys of Tropical Atlantic Forest Streams, Southeastern Brazil

Abstract

AbstractKnowledge of the adequacy of field sampling protocols is critical for detecting species and assessing biological conditions. Several studies have been conducted to determine the appropriate electrofishing distance for estimating fish assemblage richness in temperate North American streams. We tested whether electrofishing 40 times the mean wetted channel width was sufficient for estimating species richness and developing precise metrics for Atlantic Forest streams in southeastern Brazil. We sampled 32 sites with differing gradients, substrates, and anthropogenic pressures. Our results show that 40 channel widths were not sufficient to estimate species richness in those systems, presumably because of the high number of rare species. However, 40 channel widths were sufficient for applying other metrics of assemblage condition (e.g., number of common species, percentage of tolerant individuals, percentage of Characiform species, percentage of Siluriform species, percentage of water column species, percentage benthic species, Shannon diversity, dominance). This suggests that 40 channel widths are an appropriate sampling distance for applying environmental assessment protocols to Atlantic Forest streams.